AI & SOCIETY

https://doi.org/10.1007/s00146-018-0810-3

ORIGINAL ARTICLE

Skillful coping with and through technologies

Some challenges and avenues for a Dreyfus-inspired philosophy of technology

Mark Coeckelbergh1

Received: 22 May 2017 / Accepted: 24 January 2018

© The Author(s) 2018. This article is an open access publication

Abstract
Dreyfus’s work is widely known for its critique of artificial intelligence and still stands as an example of how to do excel-
lent philosophical work that is at the same time relevant to contemporary technological and scientific developments. But for
philosophers of technology, especially for those sympathetic to using Heidegger, Merleau-Ponty, and Wittgenstein as sources
of inspiration, it has much more to offer. This paper outlines Dreyfus’s account of skillful coping and critically evaluates its
potential for thinking about technology. First, it is argued that his account of skillful coping can be developed into a general
view about handling technology which gives due attention to know-how/implicit knowledge and embodiment. Then a number
of outstanding challenges are identified that are difficult to cope with if one remains entirely within the world of Dreyfus’s
writings. They concern (1) questions regarding other conceptualizations of technology and human–technology relations,
(2) issues concerning how to conceptualize the social and the relation between skill, meaning, and practices, and (3) the
question about the ethical and political implications of his view, including how virtue and skill are related. Acknowledging
some known discussions about Dreyfus’s work, but also drawing on other material and on the author’s previous writings,
the paper suggests that to address these challenges and develop the account of skillful coping into a wider scoped, Dreyfus-
inspired philosophy of technology, it could take more distance from Heidegger’s conceptions of technology and benefit from
(more) engagement with work in postphenomenology (Ihde), pragmatism (Dewey), the later Wittgenstein, and virtue ethics.

Keywords Dreyfus · Skillful coping · Technology · The social · Virtue · Wittgenstein · Dewey

1 Introduction
For philosophers of technology, Dreyfus’s work is an attrac-
tive and outstanding example of how one can combine philo-
sophical thinking with a focus on specific technologies such
as artificial intelligence (AI), and indeed how one can do
excellent philosophy by thinking about technology. Com-
pared to the often shallow contemporary debates about AI,
robotics, or transhumanism, for example, Dreyfus’s think-
ing is a relief. His philosophical arguments against the idea
that AI can give us a human-like general intelligence are
still highly relevant today, when absurd Platonic–Cartesian
ideas such as mind uploading or projects trying to artificially
* Mark Coeckelbergh
mark.coeckelbergh@univie.ac.at
1 ment’ relation as the only relevant human–technology rela-
University of Vienna, Universitätsstraße 7, 1010 Vienna,
Austria
recreate the human brain are gaining more traction than they
deserve.
But there is more to Dreyfus’s work than thinking about
AI, and there is much more to be gained from it for thinking
about technology, especially for those philosophers sympa-
thetic to using Heidegger, Merleau-Ponty, and Wittgenstein.
This paper focuses on Dreyfus’s account of skillful coping
and critically evaluates its potential for thinking about tech-
nologies today. After sketching a brief working account of
skillful coping and exploring what it means for thinking
about technology, a number of challenges are identified and
some avenues to cope with these challenges are opened up—
partly by drawing on Dreyfus and partly by going against or
beyond Dreyfus.
In particular, the paper identifies problems related to
(1) Dreyfus’s selection of tool use as the only way to con-
ceptualize technology (and what Ihde calls the ‘embodi-
tion), together with the unfortunate repetition of the (later)

13
Vol.:(0123456789)

Heidegger’s inconsistencies and bias regarding contempo-
rary technologies; (2) his underdeveloped conceptualization
of the social and of the relations between skill, meaning,
and practice; and (3) the question regarding the ethical and
political implications of his view, including the question how
skill and virtue are related. In the course of the paper, there
will be references to some known debates about Dreyfus’s
approach, which are often related to its Heideggerian dimen-
sion, such as his Heideggerian interpretation of ground and
his conceptualization of the social in terms of ‘das Man’
(translated by Dreyfus as ‘the Anyone’). But the paper will
also introduce other material to the discussion about Drey-
fus’s work, such as Ihde and Dewey.
This will lead to opening up some avenues that are not
(fully) present in Dreyfus’s own writings. In response to the
problems identified, the paper proposes some ways in which
a Dreyfus-inspired phenomenology of technology use and
skill can be constructed and elaborated, but also revised, to
address or avoid the problems previously indicated. Concep-
tual tools offered include philosophical schools and tradi-
tions Dreyfus sympathized with, but—focusing mainly on
Heidegger and Merleau-Ponty—never (fully) used: postphe-
nomenology, Dewey, the later Wittgenstein, and the tradition
of virtue ethics—the latter interpreted in a way that replaces
the Platonic and Aristotelian obsession with theory with the
love of practical wisdom and know-how and an emphasis on
the exercise of skill. In the course of the paper, some of these
suggestions and new directions are supported by referring
to the author’s own work on skill and technology, which is
inspired by Dreyfus’s view but also critical of it.
In the conclusion, the suggestions and directions offered
in the course of the paper are drawn together and presented
as elements for a Dreyfus-inspired (but perhaps not Drey-
fusian) research program in philosophy of technology. It is
also stressed that the general direction Dreyfus took in his
thinking on skill is not only important for philosophy of
technology, but also stands as one of the conceptual building
blocks all philosophers have at their disposal today to cope
with a ghost that continues to haunt both philosophy and AI:
the crazy idea that knowledge and thinking can be entirely
formalized and divorced from lived experience and active
coping in the world.
2 Dreyfus’s account of skillful coping
and its implications for thinking
about technology
2.1 Skillful coping
Dreyfus’s “skillful coping” view is based on his reading of
the early Heidegger of Being and Time (1927), and has later
found expression in the account of skill development, which
13
AI & SOCIETY
he worked out with his brother Stuart Dreyfus (1980a), and
in subsequent papers—some of which are bundled in the
book Skillful Coping (2014)—that further elaborated this
account and defended it against some objections. The main
idea is that in practical activity and skillful coping, we do
not rely on rules or mental representations, at least when
we gain a sufficiently high degree of expertise and mastery.
In Being and Time (1927), Heidegger already made a dis-
tinction between two modes of experiencing a tool: when
we use a tool, the tool withdraws, we do not notice it. It
is ready-to-hand. Under other circumstances, for instance
when something goes wrong, it might become present-at-
hand. But the “default” mode, we may say, is ready-to-hand.
For example, when I use my computer to write this paper, I
do not notice the computer; but when the operating system
crashes, the technology becomes present-at-hand. Moreo-
ver, in our skillful coping, we do not usually need mental
representations. In Phenomenology of Perception (1962),
Merleau-Ponty already emphasized embodiment and what
Dreyfus interprets as a critique of mental representation
(Dreyfus 2002a). He explains that Merleau-Ponty’s terms
‘the intentional arc’ and ‘getting a maximal grip’ imply that
skill, as a tight connection between the agent and the world,
is not stored in the form of representations in the mind, but
instead as an embodied tendency to respond (Dreyfus 2002a,
367); we are already set to respond, so to speak (373). Influ-
enced by Heidegger and Merleau-Ponty, Dreyfus, in turn,
argued that our knowledge of handling things involves a
know-how that is not explicit. We are absorbed in what we
do, perhaps even in a state of “flow”, and we respond to the
situation: ‘According to Merleau-Ponty, in absorbed, skill-
ful coping, I don’t need a mental representation of my goal.
Rather, acting is experienced as a steady flow of skillful
activity in response to one’s sense of the situation.’ (Dreyfus
2002a, 378).
Against representationalism and Cartesianism in philos-
ophy of mind and action, Dreyfus argued that the phenom-
enology of, and knowledge involved in, practical activity
and skill acquisition is not based on rules or mental repre-
sentations, but involves a coping that, especially when one
is an expert, is based on implicit know-how and intuition.
Our activity may be goal directed, but we do not think
about the goal. Deliberation only comes in when coping is
blocked (Dreyfus 2002a, 381), or indeed when we learn a
new skill. We need rules when we are novices, but experts
can do without them: rules are for beginners (Dreyfus
and Dreyfus 1980a). As we move through stages of skill
development, we rely on concrete experience rather than
abstract principles. In the paper written with his brother
Stuart they give the examples of language learning, chess
playing, and flying an airplane, which entail perfor-
mances and responses without the application of rules or
principles: ‘The expert pilot, having finally reached this
AI & SOCIETY
non-analytical stage of performance, responds intuitively
and appropriately to his current situation.’ (Dreyfus and
Dreyfus 1980, 12). Moreover, experts can have the experi-
ence that they are intensely absorbed in what they do. Here
there is no longer self-monitoring at all:
‘masterful performance only takes place when the
expert, who no longer needs principles, can cease to
pay conscious attention to his performance and can let
all the mental energy previously used in monitoring his
performance go into producing almost instantaneously
the appropriate perspective and its associated action.’
(Dreyfus and Dreyfus 1980, 14).
Other terms that qualify this experience are ‘holistic’ and
‘integrated’. Elsewhere Dreyfus argued that bodily skills
such as swimming requires a kind of unification and inte-
gration, and cannot be reduced to rules and a sequences of
movements learned by beginners:
‘Even though bodily skills, for example, are sometimes
learned by following rules which dictate a sequence
of simple movements, when the performer becomes
proficient the simple movements are left behind and a
single unified, flexible, purposive pattern of behavior
is all that remains. It makes no sense to attempt to
capture a skill by using a representation of the original
elements used by beginners, since these elements are
not integrated into the final skill.’ (Dreyfus 1980, 9).
In contrast to computers, which need to recognize an
object to manipulate it, Dreyfus argued, humans do it in a
more gestalt-like manner and instead can ‘manipulate an
object in order to recognize it’—in other words, they cope
with the object (Dreyfus 1967, 21).
Against McDowell, Dreyfus argued that experience is not
inherently conceptual and that we have to overcome the myth
of the mental (Dreyfus 2005). While it is plausible that, as
Rietveld has argued, both McDowell and Dreyfus share the
view that our engagements with the world are situated and
are each in their own way committed to phenomenology—
a view McDowell reached through Wittgenstein (Rietveld
2010, 185); Dreyfus mainly through Heidegger—Dreyfus
did not share the former’s strong emphasis on the rational
and conceptual. Perception is not conceptual “all the way
out”, as McDowell agued; instead, Dreyfus focused on non-
conceptual embodied coping (Dreyfus 2005, 47). Against
cognitivism and against representing things from ‘a detached
theoretical perspective’ (49), he argued that our everyday
coping does not require our minds to impose meaning onto
the world: ‘We need to consider the possibility that embod-
ied beings like us take as input energy from the physical
universe and process it in such a way as to open them to a
world organized in terms of their needs, interests, and bodily
capacities without their minds needing to impose a meaning
on a meaningless Given.’ (Dreyfus 2005, 49). Instead, in our
everyday coping, the world is already meaningful for us.
Moreover, whereas McDowell interprets Aristotle’s
account of practical wisdom (phronesis) in terms of respond-
ing to reasons, Dreyfus, using Heidegger, argued that prac-
tical wisdom is more like seeing what to do in a particular
situation; it is about being ‘responsive to the specific situa-
tion’ (Dreyfus 2005, 51). Or more precisely: both thinkers
share the interpretation of Aristotle’s concept of practical
wisdom in terms of responding to specific situations (Riet-
veld 2010, 189), but they have a different view of the role
of reasons and concepts in this responding. According to
McDowell, we respond to reasons and rationality ‘perme-
ates’ even unreflective action (McDowell 2007, 368) and
concepts are ‘operative’ in unreflective action (372); they
show themselves in the activity. It is rationality in action (see
also Rietveld 2010, 193). According to Dreyfus, by contrast,
reasons are ‘retroactive rationalizations’ (Dreyfus 2005,
51) and unreflective action is not about rationality. Dreyfus
repeated that when we learn a skill, we may use rules as
a kind of aid first, but like the training wheels of children
who learn to ride a bicycle, when we become experts we
can leave them out and we switch to ‘a more involved and
situated way of coping.’ (52).
Dreyfus’s point was not that mental representation, goal
directedness, reasons, etc. play no role at all in human expe-
rience; the point was that they do not usually play a role
when we decide as experts; they may play a role at earlier
stages (Dreyfus 2002b, 413). He thus stressed that some cen-
tral cases of intelligent behavior—indeed most of our eve-
ryday coping—do not require mental representations (414).
He rejected the view that we can only act if we have reasons
to do so; reasons play no role and rather get in the way.
Instead we have to sharpen our ability to make refined dis-
criminations (Dreyfus 2005, 52). And of course a game has
rules, but these rules are not stored in the mind or not even
followed; instead, the expert copes and in her coping she is
sensitive to the rules of the game (53). Everyday coping is
intentional, but not conceptual. We respond to what Gibson
called ‘affordances’, which happens without thinking at all
(56). For example, doors afford going in and out, but we do
not have to think about them—let alone calculate. When we
learned a skill, we masterfully respond to specific situations.
This gives us a familiarity with the world and openness to
the world, which is not a totality of objects or of states of
affairs. Instead we know how, and our skill opens a world.
Our conceptual capacities grow ‘out of our nonconceptual
ones’ (61).
In ‘The Return of the Myth of the Mental’ (Dreyfus
2007a), Dreyfus continued his arguments against McDow-
ell, this time against the claim that in mature human
beings, embodied coping is permeated with mindedness.
Dreyfus stressed again the non-conceptual and non-mental
13

content of skillful coping and performing. We can moni-
tor our performances while performing, but this actually
degrades the performance itself, especially since it dis-
rupts the mindless absorbed coping. Of course we can step
back and reflect, but also that kind of conceptual intention-
ality is based on a ground-floor of motor intentionality,
and there is still involved coping going on in the back-
ground. Luckily we are only ‘part-time rational animals’
(Dreyfus 2007a, 354); we can exercise our freedom as
humans not so much by thinking rationally, but rather by
opening ourselves ‘to being bound’—entering our involved
coping (355). In this coping there is no thinking subject
(358); rather, the world may afford or solicit:
‘According to Merleau-Ponty, at the most basic level
of being in the world, what does the grasping is not
the mind but the body with its nonconceptual coping
skills, and what is grasped are not unified, proposi-
tional structures that one can observe and entertain in
thought, but more or less indeterminate solicitations
to act.’ (Dreyfus 2007a, 359).
To use one of Dreyfus’s examples: when we take a
doorknob to open the door, we do not think about this
affordance or solicitation. We respond and do all this with-
out thinking. Again, we can step back, but only against
a pervasive background (363). Yet like in his previous
paper on the topic, Dreyfus acknowledges that there is
still an outstanding issue: we still want to know how the
conceptual world is related to, or emerges from, the non-
conceptual world (364). (I will return to this issue.)
Moreover, against Searle, Dreyfus argued that social
meaning and norms are not representational (Dreyfus
1999) and do not require intentional action. According
to Dreyfus, our experience of our everyday involvement
conflicts with the logic of constitution and reconstruction
proposed by Searle (Dreyfus 2001a, 181). The problem
is not that analysts construct a rational account; we can
do that. But it is wrong to suppose that we give mean-
ing to artefacts. According to Dreyfus, following Hei-
degger and other existential phenomenologists, meaning
must not be brought into a meaningless universe; there
is already a meaningful world (Dreyfus 2001a, 186). We
are ‘from the start socialized into a world in which we
cope with equipment’ (187). For example, to use money
(Searle’s paradigmatic example), there is no need to think
“this piece of paper counts as money”; we just use it and
already see it as valuable if we grew up with it (189).
There is already a meaningful world. There is already a
cultural style, which we already pick up as a baby. There is
already something binding on us, and we learn it without
having to be conscious about the rules (195–196). Else-
where Dreyfus wrote that there is a background to which
we respond: a background of practices, which does not
13
AI & SOCIETY
consist of representations at all (Dreyfus 1980b, 9). (I will
soon say more about this.)
2.2 Technology
Whether or not Dreyfus’s model of skill acquisition is a
sound interpretation of Heidegger (for a discussion see
Breivik 2007), it is an attractive model for understanding
the everyday use and handling of technologies. Technology
can be a tool we use, or (as we can add using postphenom-
enology—see the next section) it can be a medium through
which we handle objects. In both cases, following Dreyfus
we can say that skillful coping is involved—at least when
we are expert users. There is no need to think about the
technology; if we are experts there is even no need to think at
all. We respond skillfully to the situation. Through technol-
ogy the world offers affordances and solicitations, without
any need for representation by means of concepts. Many
examples Dreyfus uses are technological, such as using a
doorknob or driving a car: the expert driver does not need
to think about rules on how to shift the gears; if the driver is
sufficiently skilled, there is a bodily response appropriate to
the situation. Thus, apart from everything else it may be (e.g.
a significant contribution to epistemology, philosophy of
mind, etc.), Dreyfus’s account of skillful coping is to be seen
as a contribution to thinking about technology, to the extent
that it enables us to say more about the kind of knowledge
and experience involved in the use of everyday technologies
such as hammers, cars, and doorknobs. It seems especially
suitable to think about the handling of tools and equipment.
Sometimes Dreyfus explicitly addressed the question
how we use tools. Drawing on Heidegger, Merleau-Ponty,
and Polanyi, he argued that how we use a tool differs from
explicit experience and representational knowledge of an
object. Taking up Merleau-Ponty’s example of the blind
man’s stick, he writes:
‘A blind man who runs his hand along the stick he uses
to grope his way will be aware of its objective charac-
teristics. When he is using it, however, he is not aware
of its objective traits nor of the pressure in the palm of
his hand. Rather, the stick has become, like his body,
a transparent access to objects.’ (Dreyfus 1967, 27).
According to Dreyfus, this use has nothing to do with
making calculations or with following rules or principles.
It is an embodied kind of coping, a learning of skill: I first
happen to touch something and then I repeat it, correct it,
and so on. I consciously intervene to improve my perfor-
mance. But there is not first a knowledge of rules. Whereas
the skilled performance may be described (from the outside,
by science—for instance to build a robot) in terms of rules,
‘these rules need in no way be involved in producing the
performance.’ (Dreyfus 1967, 29). Hence—and luckily, we
AI & SOCIETY
may add—we can dwell in the world without having to for-
malize everything (31).
Does this mean that experts use no rules at all? I propose
to distinguish between a ‘strong view’ and a ‘weak view’.
According to the strong view, experts qua experts use no
rules at all when they do what they are good at. Perhaps they
use rules when they learn something new, but normally they
do not need rules. According to the weaker view, experts use
both rules and implicit knowledge, depending on the situa-
tion. Dreyfus seems to have held the strong view. For the use
of technologies, it means that expert users of technology use
it without rules and that everyday coping with and through
technology happens without the use of concepts.
This strong view also implies that even the attribution of
situation-specific aspects (in and by thinking) is excluded. In
this response to McDowell (2007b), who, as we have seen,
emphasizes the role of concepts and rationality even in
unreflective action, Dreyfus also uses the example of using
tools—now taking up Heidegger’s example of the hammer.
Dreyfus argues that we do not even need a situation-specific
concept:
‘Indeed, in our everyday coping, which he (Heidegger)
calls ‘‘pressing into possibilities”, we don’t deal with
objects with general properties like weight, nor with
situation-specific aspects like too heavy. Rather, when
everything is going well and we are absorbed in our
coping, the equipment we are using “withdraws”
(where, as we shall see, this does not mean becomes
implicit). Then there is no place for a demonstrative
concept pointing out our equipment as anything. We
do not attribute a general property or even a situation-
specific aspect to it; we just cope.’ (Dreyfus 2007b,
371–372).
Thus, according to Dreyfus, our everyday coping is not
permeated by conceptuality at all. When we look at the phe-
nomenology of tool use, we usually see absorbed coping. In
absorbed coping, when performance is at its peak, there is no
monitoring going on; there is ‘flow’ (Dreyfus 2007b, 373).
He uses again the example of a tool: ‘I don’t see the door-
knob as a doorknob when I’m absorbed in using it’ (375).
There is a direct response to the situation, no concept or
thinking. Conceptual mindedness, Dreyfus suggests, is only
possible against the background of non-conceptual, absorbed
coping (376–377).
These examples show that technology, in the form of
tools or equipment, is important for Dreyfus; it is part of
our (life) world. The already-meaningful world is a world
in which we cope with equipment: we are socialized into a
world ‘in which we cope with equipment’ (Dreyfus 2001a,
187). In this sense, Dreyfus already has a philosophy of
technology, or rather this work, the philosophy of skilled
coping, is entangled with the philosophy of technology. For
thinking about technologies today, his view has great poten-
tial. It seems suitable to be connected to other approaches
in contemporary philosophy of technology such as postphe-
nomenology (see the next section) that also emphasizes how
technology is part of the lifeworld.
However, if we want to develop his account of skilled
coping into a wider scoped philosophy of technology, we
need more reflection on technology. Not all technology is
‘tool’ or ‘equipment’, or is something we handle and which
disappears from view in our use and handling; there are
more phenomena and experiences we can associate with
technology. And, related: it is striking that Dreyfus’s exam-
ples of everyday artefacts—at least in his work on skilled
coping—are all old technologies; what does his account
mean for contemporary technologies and media such as
the internet and smartphones? Surprisingly, Dreyfus was
far more pessimistic about contemporary technologies than
one would expect from his account of skillful coping. To put
it somewhat simplistically: he did not have a problem with
hammers, but he did have a problem with the internet. Can
this discrimination be justified, or is it a bias—one not dis-
similar to the bias Heidegger had against modern technolo-
gies? And can it be justified on the basis of Dreyfus’s own
theory of skillful coping?
3 Problems with Dreyfus’s conceptions
of technology
If we want to make Dreyfus’s account of skill fruitful for
philosophy of technology, we encounter a number of prob-
lems that are related to the limitations of Dreyfus’s Heideg-
gerian conceptions of technology.
Like in Heidegger texts, there are roughly two ways in
which technology appears in Dreyfus’s work. The first one is
technology as a tool, used in everyday coping. This meaning
is based on the early Heidegger—Being and Time (1927).
The second one is technology in its modern and contempo-
rary forms, such as the internet. Technology is then seen as
a treat to meaning and skilled engagement. Here one could
see the influence of the later Heidegger, who embraced craft
work but argued that modern technology is enframing and
constitutes a danger (Heidegger 1977). Both conceptualiza-
tions of technology can be criticized, have been criticized
(for instance in postphenomenology), and must be criticized.
The first meaning of technology (technology as tool)
makes sense, but is limited to one human–technology rela-
tion. Of course we often relate to technology as a tool, and
it is the case that in our use of the tool, the tool usually
becomes invisible. But as Ihde has shown (Ihde 1990), there
are more human–technology relations. Dreyfus focuses on
what Ihde calls embodiment: a relation in which the technol-
ogy withdraws in use. But technology can also appear, for
13

example, as reality to be interpreted or “read” (hermeneutic
relation) or as quasi-other (alterity relation). Phenomeno-
logically speaking, a thermometer is not so much a tool as
a feature of the world we perceive. And if we encounter a
humanoid robot, we do not always relate to it as a tool, but
often the robot is interacted with as if it was a human other.
When it comes to his work on skill, Dreyfus unnecessar-
ily limits his discussion of technology to tools. It would be
interesting to discuss the role of skills and embodied cop-
ing in, for instance, the “reading” of technologies and in
our encounters with machines. Here postphenomenological
analysis is more sensitive to the various ways in which we
experience technologies, whereas Dreyfus has more to say
on the skill dimension. A combination of both approaches
seems promising. For example, if it is the case that many
technologies ‘mediate’ our experience and actions (Verbeek
2005 and subsequent work), then using Dreyfus enables us
to argue that they do this by re-shaping our skilled engage-
ment with the world and with others.
The second meaning of technology appears in Dreyfus’s
work on the internet (Dreyfus 2001b). Here contemporary
technology is seen as alienating and meaningless, in contrast
to craftwork (again referring to the use of tools). Accord-
ing to Dreyfus, ‘distance-apprenticeship is an oxymoron’
(Dreyfus 2001a, 67). Moreover, Dreyfus assumes that when
we ‘enter cyberspace’ we leave behind our ‘vulnerable,
embodied selves’ and avoid risk as disembodied web surf-
ers, remote from the real (6).
Dreyfus has rightly raised issues concerning distance,
embodiment, and meaning. It is important to evaluate uses
of contemporary technologies in these terms. But there are
at least the following problems with this analysis.
First, as Verbeek and others have argued, Heidegger’s
conception of modern technology is biased and conserva-
tive (Verbeek 2005). It is not clear why modern technology
would be necessarily more alienating than ancient technolo-
gies. Dreyfus seems to adopt this view. In my book Envi-
ronmental Skill (Coeckelbergh 2015) I have argued that the
internet is not necessarily alienating. Technology might
sometimes lead to more engagement with the world, for
instance when it attends us to features of our environment we
may not have noticed. For this purpose it is also necessary
to describe and reflect on the phenomenology of technolo-
gies as opposed to ‘modern technology’ as Heidegger did
(see also again Verbeek 2005). Second, to see technology
as meaningless, or as leading to less meaningful lives (or
even nihilism), contrasts with Dreyfus’s (and Heidegger’s)
own view that the world is always already meaningful and
that we use technology as embodied beings. This leads us,
third, to the questions regarding embodiment. While Dreyfus
is right to expose the problematic recurrence of Platonic
dream of leaving the body behind (and correctly argues that
Nietzsche would reject transhumanism since his vision of
13
AI & SOCIETY
transcendence implied affirming, not denying our bodily and
mortal nature), it is not clear why our use of the internet,
or for that matter all contemporary digital or information
technology, would be disembodied use. Dreyfus seemed to
have confused the Platonic and Cartesian visions of some
net enthusiasts with the phenomenology of the lifeworld.
As I have argued against Dreyfus in Human Being @ Risk
(Coeckelbergh 2013a), when we are online or in virtual
worlds we are still embodied and our so-called “online”
actions and thinking is still embodied: ‘When we are online,
we don’t leave our body at home.’ (Coeckelbergh 2013a,
130). Use of these technologies does not close off possi-
bilities for skilled engagement (Coeckelbergh 2013a, 132).
Elsewhere, I have also argued, against Dreyfus, that when
we use the internet, we are actively related to our environ-
ment as embodied beings, and information technology does
involve skills (Coeckelbergh 2011a, 152). In other words, it
is also a form of skilled coping. And, in principle, it could
also be a site for the acquisition of new skills and the learn-
ing of mastery. It is true that, as Dreyfus argued, for masters
to respond immediately to present situations in a masterful
way, they must have had experience of embodied successes
and failures (Dreyfus 2001a, b, 67). But use of the inter-
net, including distance learning, is also a kind of embodied
learning—albeit (it must be admitted) of a different kind.
Dreyfus may have contributed to dispel the myth of
the mental; he still seemed to have believed in what we
could call the myth of the virtual (or the myth of the real
as opposed to the virtual): the belief that somehow, when
we use the internet and related contemporary information
technologies, we are in a different reality (an online, vir-
tual world), which is disconnected with the real reality. But
Heidegger’s and Dreyfus’s own approaches try to move
beyond dualist frameworks, and on the basis of Dreyfus’s
own account of skillful coping, one could claim that in
these uses of contemporary technologies, the technology
also often withdraws, and that embodied coping is going
on, mediated by the new technologies. There is no outside
to being-in-the-world and existence; we remain earthly and
situated beings. There is one lifeworld (Coeckelbergh 2011a,
153). If there is alienation, it cannot be defined in this way.
If there is alienation at all, it needs to be conceptualized as a
problem of a particular use and relation, rather than as some-
thing linked to the technology as such. There is a problem
with regard to distance and telepresence, but it should not
be conceptualized in terms of a total absence of embodi-
ment. At most, it is a different kind of embodiment. There
might be a problem with regard to meaning, but it cannot be
the total loss of meaning, since if that use is a skillful cop-
ing it has meaning, and in the spirit of Dreyfus, Heidegger,
and the later Wittgenstein, we can say that use takes place
against the background of a meaningful whole. It may be
the case that the processing of date itself is meaningless; but
AI & SOCIETY
phenomenologically there is no “in itself”; our experience
and our engagement with computers our phones and our
coping with data never is without meaning, and makes sense
against a wider background of meaning.
This takes us to Dreyfus’s view of ‘the background’,
which I shall further discuss in the next section.
4 Problems with Dreyfus’s (under-)
conceptualization of the social
A related but different outstanding problem raised by Drey-
fus’s account of skillful coping is how skillful coping is
related to meanings not limited to a particular skilled per-
formance, meanings that have to do with the relevant prac-
tice, and more generally the social. Skillful coping does not
happen in a social vacuum, but is always already linked to
larger practices and to a wider social background. Although
Dreyfus acknowledges this, the account of skillful coping
itself is not very helpful for conceptualizing it. If we want an
approach to philosophy of technology that says more about
technology than what happens at the level of individual cop-
ing, and that is still inspired by Dreyfus, then his underde-
veloped dimension of Dreyfus’s work needs to be exposed,
discussed, and repaired.
Partly Dreyfus’s problem here is also a problem for phe-
nomenology. One recurring problem with all phenomeno-
logical approaches—including postphenomenology—is a
tendency to limit phenomenological analysis to individual
subjects and their relation to the world. This is also applica-
ble to other phenomenological approaches in philosophy of
technology. For example, in Ihde (1990) and Verbeek (2005),
the social and the cultural is of course acknowledged, but
there is a focus on an “I” that relates to “world”, mediated
by technologies. The root of this problem lies in the classic
phenomenological tradition (Husserl and Heidegger), or at
least in individualistic readings of it. From the perspective
of other approaches in philosophy of technology such as
critical theory and (I will add) pragmatism, but also from
approaches in the social sciences (social studies of science
and technology, for instance), it must be asked if Dreyfus
sufficiently conceptualized the social and took it seriously.
For philosophy of technology, this is an important criticism
to discuss, since, as social studies of science and technology
approaches have always emphasized, technology is embed-
ded in a social context and cannot be appropriately studied
without taking into consideration the social dimension.
Consider Dreyfus’s example of driving a car. The account
of skillful coping is very good in describing the operation
of the car in the sense of changing gears, braking, steer-
ing—and indeed the coordination of all these operations.
When one is an expert driver, the skills involved no longer
require instruction or rules; they are embodied and intuitive.
But driving expertise is not only about the operations as
such; it also consists of participating in traffic. Traffic is a
social phenomenon and hence driving, in the wider sense,
is a social practice. To be skilled in driving means not only
to be able to operate the machine but also to respond to
situations in traffic—including its morally relevant features
(Coeckelbergh 2016). But describing this requires a phe-
nomenology that is not only about how I related to “world”,
but also about others. It requires a more social and relational
phenomenology.
Let me first explore how this link to the social may be
conceptualized from a Dreyfusian (that is, mainly Heideg-
gerian) perspective—before suggesting some alternatives
that go beyond Dreyfus and take inspiration from Wittgen-
stein and Dewey. Like Heidegger, Dreyfus does not address
the social as his primary focus. But, like Heidegger, that
does not mean that he does not mention it or that we cannot
(re)construct some conception of the social based on this
writings. A good place to start is Dreyfus’s references to
practices and to the ‘background’. Then I will turn to his
use of the Heideggerian concept of ‘das Man’ (the Anyone).
4.1 Dreyfus on the (social?) background
The question regarding the background arises from the (tran-
scendental) question about what must be presupposed for
a particular practical experience and skillful coping. For
example, driving is only possible, and makes sense, against
a background of a shared practice and shared norms and
beliefs about what counts as good behavior. Driving is also
related to other practices, such as (getting to) work and shop-
ping. Ultimately, driving is also part of an entire culture and
way of doing, which I would call, with Wittgenstein, a form
of life. (See below.) Yet “norms” and “beliefs” may already
be a bridge too far for Dreyfus, who sees background as
something that cannot be made explicit at all. Why?
According to Dreyfus, understanding must not be seen as
a merely epistemological or theoretical problem. Of course
we interpret and explicate, but the basis is practical under-
standing: everyday coping. Practical understanding must be
viewed in a holistic way. This approach is what Dreyfus
calls practical holism. He uses one of his favorite activities
to illustrate what this means:
‘Practical understanding is holistic in an entirely dif-
ferent way from theoretical understanding. Although
practical understanding—everyday coping with
things and people—involves explicit beliefs and
hypotheses, these can only be meaningful in spe-
cific contexts and against a background of shared
practices. And just as we can learn to swim without
consciously or unconsciously acquiring a theory of
swimming, we acquire these social background prac-
13
 AI & SOCIETY

tices by being brought up in them, not by forming involved in the learning of skill and in coping—is also
beliefs and learning rules.’ (Dreyfus 1980b, 7). a seat of meaning and a concrete incorporation of the
background.
Here Dreyfus seems to acknowledge the social and the
Dreyfus is right to point to the background as a seat of
cultural. Based on Dreyfus, we could see the social in
meaning that is already there, and which usually does not
terms of (the totality of) shared practices, which form the
need to be made explicit to function. But based on this analy-
background of everyday coping and beliefs, and which
sis of background, it is not so clear how the social comes in,
‘make us who we are’ (Dreyfus 1980b, 10). The social
and if the social as background must always remain entirely
and the culture are thus embodied in our practices. This
inexplicit. Does Dreyfus really mean that the background
is also the case for scientific practices, which are only
cannot even partly be made explicit in terms of rules or
possible against background practices (16). However,
norms? Is the idea of a social science, which arguably has
Dreyfus refuses to define those background practices in
this project of explicitation at its very core, entirely mis-
terms of rules. In contrast to what Quine and Davidson
taken? Is it senseless to speak of social rules or norms, for
supposed, for Dreyfus this cultural background cannot be
example, as social scientists do? And are philosophers who
made explicit and is not “in our minds” in the form of
do so entirely misguided? It is not clear why Dreyfus takes
beliefs or background assumptions which could, in theory,
on board Wittgenstein’s anti-mentalist epistemology and
be made explicit:
philosophy of mind, but does not seriously engage with
‘Heidegger, Merleau-Ponty, and Wittgenstein sug- Wittgenstein’s ideas about games and rules—ideas which
gest that this inherited background of practices can- influenced social scientists. (I will say more about a Witt-
not be spelled out in a theory because (1) the back- gensteinian route below.) Dreyfus seems to miss an oppor-
ground is so pervasive that we cannot make it an tunity here to bring in these aspects of the social. For think-
object of analysis, and (2) the practices involve skill.’ ing about technology, we seem to be stuck with an account
(Dreyfus 1980b, 7). of skillful coping that is holistic and related to ‘the back-
ground’, but that background cannot be further made explicit
What makes up the background is not beliefs, but
at all and is bound to remain mysterious.
‘habits and customs, embodied in the sort of subtle skills
An interesting exception occurs when Dreyfus does speak
which we exhibit in our everyday interaction with things
of norms in his comment on style in his response to Searle,
and people.’ (Dreyfus 1980b, 8). These cannot be reduced
already mentioned before. He suggests that there are already
to propositional knowledge or rules. We respond to what
social norms, there is already something binding on us, and
situations require.
we learn that from early ages on—without having to rely on
Increasingly, Dreyfus speaks of ‘the background’ rather
explicit rules. This makes sense and opens up the possibility
than a background of shared practices. This use of the term
of a kind of middle position, which recognizes that there are
as a noun in subject position is influenced by Searle, who
rules, norms, etc. that have some normative power over us,
speaks of ‘Background’ (Andler 2000). But, considered
but at the same time acknowledges that we do not (always?)
together with his position saying we cannot spell it out,
learn them in an explicit way and that we do not need them
this renders the background somewhat mysterious. Drey-
in an explicit and formalized form in order to cope. But this
fus argues that the background is ‘hidden and holistic’
thought is not made explicit, let alone elaborated. Other-
(Dreyfus 2012). For Dreyfus, influenced by Heidegger
wise, we must conclude that Dreyfus simply held on to the
and Merleau-Ponty, we cannot describe the background,
more extreme position that the background cannot be made
since when the world functions as background it ‘has’ to
explicit at all.
withdraw to do its job. It enables other things to show
To conclude: the social and cultural can be connected to
up (Dreyfus 2012). This is, according to Dreyfus, how
Dreyfus’s thinking about background, but this connection
the phenomenology of skilled coping works. Whereas for
is not further theorized or developed. There is, however,
Searle the background is a set of mental capacities that
one way in which he theorized the social explicitly: a very
makes possible representation, for Dreyfus meaning is
Heideggerian way, to which we now turn.
already there. As said, in his view there is no need to bring
it in via intentional acts. The world is already meaningful
4.2 Dreyfus on ‘das Man’ (the ‘Anyone’)
and the body also confers meaning. Following Heidegger,
Dreyfus holds that—as Andler put it—‘the background
Dreyfus does say something about the social in his analysis
is the condition of possibility for there to be anything at
of the early Heidegger. In a very interesting article on “the”
all of any significance to a being.’ (Andler 2000). In this
world versus “my” world (Dreyfus 1975), Dreyfus argued
sense, it is like Wittgenstein’s ‘form of life’. (I will say
that Heidegger’s analytic of Dasein was meant to overcome
more below.) And following Merleau-Ponty, the body—as
the dilemma of having to choose between prioritizing a

13
AI & SOCIETY
transcendental subject and prioritizing a social world. The
everyday world is already public; there are already goals
and equipment available in society, there are already ‘roles’
(Dreyfus 1975, 121–122). This thought makes a promising
opening to a more social analysis. But then it turns out that
Dreyfus restricts the social to a very specific conceptualiza-
tion of it in the early Heidegger: he points to Heidegger’s
“das Man” (he translates it as the ‘Anyone’). We do what
“one” (Dreyfus: “anyone”) does. This threatens authentic-
ity. But it is not just negative. Explaining a hammer means
to show what “one” does with it (124). In Wittgensteinian
terms, there is already agreement in form of life (125). Hei-
degger, however, stressed the contrast the ‘Anyone’ with
authenticity. We lose ourselves. Dreyfus is maybe more
optimistic and argues that we can still give individual style
to the modes provided by the ‘Anyone’ (129). But, neverthe-
less, we are constituted by the ‘Anyone’:
‘Dasein, which as Being-in-the-world is always a way
of acting, can and must give its individual style to
the modes of behavior provided by the Anyone. Thus
Dasein is both constituted by the Anyone in that all
significance is a social achievement, and yet self-con-
stituting by taking over significances from the Anyone
to define itself or give itself meaning. With this analy-
sis of the parasitical relation of the individual to the
social world, Heidegger gives an account of the rela-
tion of the personal to the public which (dis)solves the
Husserlian problems arising from the priority of my
world to the world.’ (Dreyfus 1975, 129).
This could be seen as pointing to a kind of (Heideg-
gerian) social philosophy. Heidegger can be interpreted as
arguing that we already find ourselves in a social world, and
that we are always socially constituted. And it is interest-
ing that Dreyfus says something about the relation between
Dasein and Anyone, which could be seen as asking a cen-
tral problem of social philosophy, but then formulated in a
Heideggerian way. However, first, even within a Heideg-
gerian framework it can and has been objected that Dreyfus
leaves out what Heidegger calls Mitsein. Olafson has argued,
against Dreyfus, that he gives a far too large role to ‘Das
Man’, which is only a deformation of our social being (Mit-
sein) (Olafson 2008). Carman, in turn, has accused Olafson
of an over-individuated notion of Dasein (Carman 2008).
Whatever the best interpretation of Heidegger may be, it
turns out that Dreyfus (like Heidegger) has again said too
little about the social, and has left too much room for inter-
pretation. His suggestion that human beings depend on the
social world is important but underdeveloped. Second, the
social is mainly seen as a given, and as one monolithic kind
of thing. The rules or the antagonistic tendencies and vari-
ations within the social are not visible, except in the form
of style. But this thought is not elaborated, and the Anyone
itself remains one undifferentiated “block” or “box”. Drey-
fus’s thoughts here are too vulnerable to the objection that
in this view there is one ‘Anyone’, which seems not up to
change or remains black boxed. Again, this is unhelpful if
we want an approach to thinking about technology that takes
seriously the social embeddedness of technology.
To remedy this, one could resort to Marx or Hegel, who
analyzed the social in terms of struggle; this may help to
open the black box, so to speak. Or one could go directly
to contemporary social sciences, which have more to say
on style and the social, and which have also a lot to say on
technology and the social [social studies of science and tech-
nology (STS), see below]. But as already suggested, there is
a thinker closer to Dreyfus who also may be of help to say
more about the social: Wittgenstein.
4.3 Wittgenstein
Wittgenstein’s post-Cartesian, anti-mentalist view in On
Certainty (1969) and his rejection of his earlier Tractatus
view that the world is a collection of fact, is in line with
Dreyfus’s view, and supports his arguments against classic
AI and his focus on how we cope with everyday problems.
The later Wittgenstein understood knowledge not in terms
of static and inner mental states or objective facts, but sug-
gested a more practical, pragmatic, process-oriented, tacit,
and embodied conception of knowledge. Whether in scien-
tific research or in everyday life, our doubts rest on a basis
of things that are not doubted, things that are accepted and
presupposed. This approach is interesting for philosophers
of technology as it may help us to discuss the issue of trust
in technology, and supports a view of technology as a skilled
and embodied practice and technique that crucially involves
implicit knowledge (Coeckelbergh and Funk 2018). But we
can also use Wittgenstein for the purpose of developing a
view of the relation between a specific practical activity (or
coping) and the wider, perhaps implicit social background.
Dreyfus has not done this. Although he sometimes referred
to Wittgenstein, he did not really use him to further elabo-
rate the holistic relation between everyday coping and the
background. But we can. Let me explore some possibilities.
One way to start such a project is to draw on scholarship
about Dreyfus that makes connections to Wittgenstein. For
instance, Rietveld’s work on the differences between Drey-
fus and McDowell (Rietveld 2010) at least suggests that it
can be useful to discuss interpretations of Wittgenstein in
order to further elaborate a Dreyfusian approach. The same
can be said about McManus’s comparison of Wittgenstein-
ian and Heideggerian ideas when discussing the concept of
“background” in Dreyfus (McManus 2007). These critical
discussions could be used as a starting point to bring more
Wittgenstein into Dreyfus.
13

Another way is to draw directly on Wittgenstein, as I have
done in my recent work on technology inspired by Witt-
genstein (Coeckelbergh 2017a, b; Coeckelbergh and Funk
2018) and in line with some suggestions in Winner (1986).
Based on the Philosophical Investigations (Wittgenstein
1953/2009), I have argued for a holistic and transcenden-
tal approach to technology which sees technological arti-
fact, in their use and in our performances, as embedded
in what Wittgenstein calls ‘games’ and ‘form of life’. The
meanings in and of the particular use and performance are
related to our activities, games, and form of life, which are
also social and cultural, and which must be presupposed
for the particular use and activity to make sense. This is a
“background”, if you wish, but one that is more practical
and less mysterious than Dreyfus’s. It is transcendental in
the sense that it is a condition of possibility for technol-
ogy uses and performances, but it is not transcendental in
a Kantian (abstract categories) or Heideggerian (a mysteri-
ous ontology of being) way. Instead, the view I proposed
connects a particular use and performance (and hence the
skill and know-how related to it) to a wider background
of games and forms of life which already contain practi-
cal know-how. Moreover, this know-how is not necessarily
implicit. It contains a lot of implicit knowledge, for sure, but
at least part of it can be made explicit in terms of rules. This
understanding of the social renders it possible for the social
sciences to say something meaningful about the social and
our social practices. But more importantly for the purpose
of using Dreyfus’s account of skillful coping: it presents a
more elaborated and arguably more balanced theory of the
relation between skillful coping and its social background,
which can at least partly be made explicit. Usually we do
not think about the rules of our social games. Usually we do
not make explicit our form of life. And maybe we can never
make it fully explicit. Maybe not everything can be captured
in rules. But surely it makes sense to sometimes formalize
for practical purposes, that is, as a tool. (See also my use of
Dewey below.) Furthermore, this view also acknowledges
the normative dimension of the background—even if these
norms may not always be visible or even if they can only be
partly explicitated. Finally, with the concept of games we
could also add that the background is not entirely given: we
can also change it. We can change our games, and partly that
can happen by changing the rules. But it can also happen
differently. Technology, for instance, can be a game changer.
It can disrupt and transform our background understand-
ings and practices. However, I share Dreyfus’s Heideggerian
intuitions in so far as I believe that such changes are small
and slow, similar to the changes in river beds—to use a Witt-
gensteinian metaphor (Coeckelbergh 2017b).
Maybe this approach also helps us to relate the concep-
tual to the non-conceptual, the challenge Dreyfus formulated
(Dreyfus 2007a): if we see the conceptual, logic, reasons,
13
AI & SOCIETY
etc. not as things that are “out there” in the world (as
McDowell and Dreyfus seem to assume) but as something
that only “lives” embodied in performance and a use, that
only exists in our engagement with the world—in particu-
lar, our use of words and a performance with words, which
itself is, like all performances, related and made possible
by an implicit background of practices—then that seems
to me at least a preliminary answer to Dreyfus’s question.
The conceptual does not have an existence outside our uses
and performances, and the use of concepts itself depends on
embodiment and on an often inexplicit background: back-
ground knowledge, experience, and practices, that are always
already social. This answer is in the spirit of Dreyfus, but—
influenced by Wittgenstein—gives more importance to the
concept of use, acknowledges a role for the conceptual, and
attends to use of language, which is often neglected in con-
temporary philosophy of technology (Coeckelbergh 2017c)
and indeed in Dreyfus’s account of skilled coping.
Another route, which is also not necessarily distant from
Wittgenstein, is STS and, earlier, sociology of knowledge,
which usually sees knowledge as socially constructed. In the
1990s, Harry Collins already criticized Dreyfus for not tak-
ing on board Wittgenstein’s notion of form of life and, more
generally, for not seeing that the structure of knowledge is to
be found in the social collectivity (Collins 1992). He writes:
‘For sociologists of scientific knowledge, most of
whom think of themselves as jumping off from Witt-
genstein’s later philosophy, it is surprising that a phi-
losopher who leans so heavily on Wittgenstein should
even have a concept of a knowledge domain apart from
a form-of-life.’ (Collins 1992, 727).
There is domain knowledge, but this knowledge has a
history—a social history in which things could have been
different. This form of life shapes my usage. For instance,
the English language or arithmetic can be seen as domains
of knowledge, but they also kind of compel me to speak in
certain ways. But at the same time that language is made
in society. Collins thinks that we have more control over
it than Dreyfus suggests. According to him, Dreyfus failed
to distinguish between individual knowledge and collective
knowledge and focused only on individual understanding
(Collins 2013). Taking up the example of car driving again:
Dreyfus talks about how we learn to drive in the sense of
operating the car, but this leaves out the knowledge we need
about traffic. This is Collins’ take:
‘Dreyfus has no concept of collective understand-
ing that is separate from the concept of individual
understanding; this is why he makes no clear distinc-
tion between expert car driving (gear changing etc.
which requires only the context of the machine and
its relation- ship to the road to be internalised) and
AI & SOCIETY

expert car driving in traffic, which requires national we trust them, and what if they assume power on the basis
driving conventions to be understood and updated of intuitions?
through immersion in the collectivity. Gear chang- The assumption made by the authors that trust necessarily
ing, like Dreyfus’s iconic example of an expertise, relies on explicit knowledge and the suggestion that Drey-
that of the chess grandmaster, does not depend on fus’s model of skill acquisition necessarily leaves out the
collective tacit knowledge, whereas driving in traffic, role of the coach are problematic. Trust does not require that
and most other expertises, does.’ (Collins 2013, 411). everything is made explicit; on the contrary, otherwise there
is no need for trust. Trust is not necessarily the outcome of
While I disagree with Collins that it is so easy to dis-
a rational process but, according to a ‘social–phenomeno-
tinguish between the two kinds of knowledge, or indeed
logical’ view, is already embedded in the social and in rela-
that one should make this distinction between individual
tions (Coeckelbergh 2011b). I also do not see why Dreyfus’s
and collective knowledge (I already mentioned Dreyfus’s
model of learning excludes the possibility of coaching. For
interpretation of Heidegger that tries to go beyond this
example, a coach can be helpful for beginners, and even for
dichotomy and with my interpretation of Wittgenstein I
experts, since a coach does not necessarily only give rules
show how use and the social are entangled), if the point is
and instructions; she might also encourage a trial and error
that driving is also about collective knowledge then this
kind of process of the kind that Dreyfus would approve of.
makes sense and helpfully brings in a social perspective.
And Selinger and Crease speak about a very specific kind of
For thinking about technology, it means that technologies
expert than the everyday expert of Dreyfus: not the expert
need to be understood not only in terms of “micro”-skill-
driver or swimmer but, for instance, an expert who is asked
ful coping; that skillful coping needs to be related to a
about health or climate change, who is socially recognized
“macro” wider social background knowledge. One could
as “expert” in the sense of a doctor or scientist, for example,
also draw on work in contemporary STS (Wiebe Bijker and
and has a specific role to inform policy as an “authority”,
others) that shows how the design of technologies is the
who has an “audience”, and so on. But their point about
outcome of a social process, to support this point.
embeddedness is important since it points again to a lacuna
More generally, on the one hand, we can learn from
with regard to the social dimension of Dreyfus’s model.
the social sciences that there is something like collective-
Focused on embodied coping, the account of skilled cop-
social knowledge, that the background can at least partly
ing and expertise seems less sensitive to social, cultural,
be made explicit and is at least partly something which we
and political context than it could be. This includes the
can influence. ‘Das Man’ is not simply given. Influenced
point suggested by the authors that experts are and should
by Heidegger, Dreyfus makes the background more myste-
be vulnerable in the sense that they can be challenged with
rious and untouchable than it is. On the other hand, we can
regard to their expertise and that they can be asked to give
learn from Dreyfus that the social cannot be made entirely
reasons—even if not everything can be explained and some-
explicit in terms of rules and beliefs, and that there are
times experts just have to say (not “admit”, this sounds too
limits to the extent to which we can control it.
negative) that they rely on their experience and intuitions.
Another relevant objection to Dreyfus, which also links
But it makes sense to say that the social and the political also
to STS and the social sciences, has been made by Selinger
includes giving reasons and deliberation. This is a lacuna in
and Crease (2003). Here the target is Dreyfus’s account of
Dreyfus’s work. And perhaps the more fundamental problem
expertise. While the authors praise Dreyfus for bringing in
is that the entire social and political playing field, in which
a first person perspective to studying expertise (rather than
experts are situated and which plays a role in what they say
a third person perspective, as the social sciences usually
and become, remains out of sight.
do), and for calling attention to the relation between exper-
How can this be conceptualized? Next to STS and Witt-
tise and the body, they complain that his account is not
genstein, pragmatism is another source which can help us to
sufficiently sensitive to the cultural embeddedness of the
conceptualize the social dimension of skillful coping, and
expert, including prejudices, ideologies, and hidden agen-
which has a lot more to say on deliberation and public dis-
das the expert may start from. Experts are not embodied
cussion than can be done within a Heideggerian framework.
and situated subjects, as Dreyfus shows us; they are also
people that are culturally and socially embedded, and since
4.4 Dewey
that gives them prejudices, agendas, and so on, we should
not naively trust them. Society might even be endangered
A thinker who is curiously absent in most discussions by and
by experts—an issue that cannot arise in Dreyfus’s account
about Dreyfus is Dewey. (An exception is a Dreyfus paper on
(Selinger and Crease 2003, 262–263). In particular, Drey-
ethical expertise—see Sect. 5.2). This is curious since there
fus’s point that experts might not be able to proposition-
are at least the following points of connection. First, like
ally justify decisions (270) seems dangerous—how can
Dreyfus and Wittgenstein, Dewey was skeptical about the

13

language of the “mental” and sees mind as embodied. But
he also understood it as emerging from the interaction in a
social context. In Experience and Nature (Dewey 1929) he
sees knowledge as ‘a mode of interaction’ (435) and argues
that meaning is not a ‘psychic existence’ (179). Dreyfus,
focused on the handling of tools, ignores this more social-
interactive view. But if we see technology as an instrument
of social cooperation, as I proposed in my interpretation of
Dewey’s view of language (Coeckelbergh 2017a, 33–37),
then this puts tool use as skillful coping in a more social
context. Moreover, whereas Dreyfus seems to put language
in a separate category (the conceptual, the symbolical, etc.)
divorced from embodied coping, for Dewey language is
both embodied and social. It is about organized interaction
with other living creatures (258). Second, Dreyfus could
have used Dewey’s conception of habit. In Human Nature
and Conduct (1922), Dewey argues that we know how by
means of our habits (Dewey 1922, 177). This idea seems to
fit Dreyfus’s account, and could help to understand our use
of technology from a more social angle. Dreyfus talks about
habit, but does not use Dewey to elaborate this social aspect
of skillful coping. That coping and those skills can be under-
stood as being part of habitual and shared ways of doing,
and our tools become tools in the context of social groups
(186). Third, Dewey’s work also seems particularly relevant
when it comes to conceptualize public discussions about
technology. In Dreyfus’s work, the public dimension is men-
tioned but not theorized. Dewey could be used to elaborate
this aspect of the social in a non-Hegelian way. But Dreyfus
rejects this route. Here a significant barrier to using Dewey
is that whereas Dewey stresses deliberation to solve social
problems, for Dreyfus a lot is going on without deliberation.
Moreover, Dreyfus seems to divorce mind and knowledge
from the social. In his response to Collins (Dreyfus 1992,
724), Dreyfus argues that not all intelligence is social; this
goes against the pragmatist view that intelligence is social
and, again, is about solving social problems. (More below.)
Finally, one could ask if Dreyfus’s account of skilled cop-
ing is sufficiently appreciative of the social-linguistic and
communicative dimension of coping. In my interpretation
of Dewey, language must be seen as a social tool (Coeck-
elbergh 2017a, 35). Perhaps language could be integrated
in a Dreyfusian–Wittgensteinian view by saying that next
to the handling of things, there is also skillful coping with
words, and both kinds of skillful coping are always embed-
ded in a social–practical context in which there is inexplicit
knowledge but also language and language games, consti-
tuting a form of life which is given and shapes our concrete
coping-performances. And here, too, one could add that in
the use of words, in coping using words, that language with-
draws, is not always visible. Indeed, it is usually so invisible
that Dreyfus managed to leave it out of his account of skill-
ful coping. But this “default” invisibility does not justify
13
AI & SOCIETY
excluding it from a theory about the kind of knowledge and
experience involved in skilled coping.
But there are also possibilities next to using Dewey, that
are certainly not far away at all from the thinking of Hei-
degger and Merleau-Ponty, but that are, nevertheless, more
appreciative of the social aspect of skill than Dreyfus was.
For example, Borgmann, a philosopher of technology who
in his praise of skilled activity is very close to Dreyfus,
manages to connect skilled activity to the social in a very
straightforward, less Heideggerian or mysterious way. He
writes:
‘Physical engagement is not simply physical contact,
but the experience of the world through the manifold
sensibility of the body. Skill is intensive and refined
world engagement. Skill, in turn, is bound up with
social engagement. It molds the person and gives the
person character.’ (Borgmann 1984, p. 42).
Borgmann (1984) stresses that when we are engaged in
skilled activity, we do not only engage with things but also
with others, for example, when keeping a stove going cent-
ers the family. A similar view can be found in Crawford’s
analysis of craftsmanship, which is about working together
and sharing a concept of good (Crawford 2009, 181). One
could also argue that skilled activity helps us to shape our
character. This takes us to the questions regarding virtue,
and more generally the ethical and political implications of
Dreyfus’s position.
5 Problems with the ethical and political
implications and the question
regarding virtue
What are the ethical and political implications of Dreyfus’s
view of skilled coping? While his account is mainly descrip-
tive and aimed at understanding the kind of knowledge
involved in skilled coping, it has normative implications.
The problem is again that it is not entirely clear what these
implications are based on Dreyfus’s own writings; but these
implications need and deserve to be further articulated and
developed, also with a view to arrive at a better thinking
about technology.
Like Reynolds (2006) I believe that if we look at his
account of skill acquisition, Dreyfus’s view is one in which a
kind of ethical comportment is more important than sophisti-
cated reasoning: moral maturity ‘is primarily about an ethi-
cal comportment to situations in the world rather than about
coming to have more sophisticated cognitions and judgments
about principles and rule-following’ (Reynolds 2006, 545).
Ethics seems to require the development of a practical wis-
dom which can respond intuitively and appropriately to spe-
cific situations. This sounds like virtue ethics. Dreyfus might
AI & SOCIETY
have been aware of this connection: when I had a chance to
talk to him in Berkeley, years ago, he suggested to me that
virtue ethics may be a good way forward when it comes to
elaborating the ethical implications of his work. But more
work is needed on this: how, exactly, does his account of
skillful coping relate to virtue ethics? It seems to me that his
account lends support to virtue ethics, albeit a virtue ethics
of a particular kind: one that does not necessarily involve
reasoning and judgment, but rather a knowing-how to best
respond to the world and to others in particular situations.
Let me develop this thought about a Dreyfusian virtue eth-
ics by engaging with some of the literature on virtue and
skill. I will end by raising the question regarding political
implications.
5.1 Virtue and skill
What does it mean to say that virtue is about skill? Annas
famously argued that virtue is like skill (Annas 1995). The
point is that there is an analogy in structure between the two.
Now, whether or not this is a sound interpretation of Aristo-
tle (for a discussion see Stichter 2007), this is an interesting
view in the light of Dreyfus’s emphasis on skill. Annas’s
claim concerns intellectual virtue; her project was to under-
stand intellectual virtue. But what if virtue tout court is like
skill? And what if virtue is not so much about intellect at
all? Indeed, perhaps there is more than an analogy: in the
spirit of Dreyfus one could argue that this is still a far too
intellectualist conception of virtue. For Annas, virtue as skill
requires reflection (Annas 1993) and the development of an
intellectual component (Annas 1995). For her, the stress is
on reflection and decision: ‘The skill analogy requires that
the agent reflect and achieve by reflection a unified grasp
of the general principles underlying her patterns of action
and decision.’ (Annas 1993, 67–68). But is this reflection
and decision necessary and sufficient for virtue? Stichter
convincingly argues they are not necessary:
‘A person’s ability to explain herself can be less than
the person’s ability to know how to act in the situation.
The intellectual requirements that Annas discusses are
relevant to any social discourse we have about moral-
ity, but they are not necessary for achieving expertise.’
(Stichter 2007, 194).
Annas admits that many skills are not like the skills she
describes, but then still endorses her intellectualist concep-
tion of skill, going against our intuitions and experience:
indeed she defends an account of skill that ‘does not fit
numerous examples of actual skills’ (Stichter 2007, 187).
She thus perpetuates an intellectualist tradition of think-
ing about skill that started at least with Plato and Aristotle.
It ignores recent thinking about skill including Dreyfus’s
work—which was already available when Annas wrote. The
point is not that I disagree with Annas’s interpretation of
Aristotle: he may well have an intellectualist, theory-ori-
ented understanding of virtue, skill, and practical wisdom
(phronesis); this is at least also my impression when I read
him. As McPherson puts it: ‘Aristotle himself, it is worth
recalling here, classified phronesis as an intellectual virtue,
aiming at truth, rather than as a practical virtue, aiming at
goodness.’ (McPherson 2005, 706). But if this is right, there
is no need to follow Aristotle on this. Learning from Drey-
fus, we can do better and develop theory that is closer to
the actual phenomena of skilled coping and their relation to
virtue and good.
Indeed, as becomes clear on the basis of my summary
of Dreyfus’s account, and as Stichter also shows (2007,
192), Annas’s view goes against a Dreyfusian way of think-
ing about skill: for Dreyfus, skillful coping does not at all
require reflection. Hence, if we use skill as a model for vir-
tue and use Dreyfus’s view of skill, then virtue would also
be virtue without an intellectual component. It suffices that
we know how to do good in the sense that we know how to
do good in practice. Theoretical knowledge or reflection,
then, are neither necessary nor sufficient for virtue, and can
even get in the way of this virtue-as-skilled-engagement and
virtue-as-performance. A standing back, so it seems, is then
not required. For Annas, this would not be genuine skill.
According to her, virtue is about having a disposition to act
for reasons (Annas 2007). But, from a Dreyfusian point of
view, there is no good reason why we have to follow her
conception of virtue. Thus, let us go further than Annas, and
argue that virtue is not only analogous to skill but entirely
a matter of skill, but then understood as embodied coping,
and as not needing anything else than know-how: knowing
how to do good (in practice).
This is a conception of virtue I defended in my own think-
ing about technology, skill, and virtue. If we think about
ethics of technology as a virtue ethics, if we define virtue in
terms of skill, and indeed if we define technology in terms
of skill (Coeckelbergh 2012), then we get the following con-
ception of virtue: virtue is a matter of practical wisdom,
which implies a knowing-how and skill. Being “wise” is
etymologically related to a “way” (Coeckelbergh 2015, 118
footnote 2). (Here one could also consider Eastern concep-
tions of virtue, for instance in Daoism and Confucianism.)
Influenced by Dreyfus and Dewey, I have argued that virtue
is about knowing how to do things. Virtue is not about hav-
ing particular properties, but about practicing virtue. There
is not first a ‘logos’ which then needs to be applied to the
real world. Rather, virtue is about having ‘ethical know-how’
and doing it, practicing it (Coeckelbergh 2011a, 162). And
if we then see technology in terms of skill, we have a view
in which virtue, skill, and technology are integrated. There
are no longer two sides—reason and intellect versus embod-
ied coping and material technologies—but there is only one
13
 AI & SOCIETY

skillful coping, engagement, and practice, and virtue is all sions are necessarily crude since they have not been
about the quality of that coping, engagement, and practice. refined by the experience of the results of a variety
For example, environmental virtue is not something abstract of intuitive responses to emotion-laden situations and
and external (say, a principle) that is imposed on the life- the learning that comes from subsequent satisfaction
world as it were from outside, but rather something that goes and regret. Therefore, in familiar but problematic situ-
on in the lifeworld and that transforms it. It is about how we ations, rather than standing back and applying abstract
(skillfully and technologically) relate to our environment. principles, the expert deliberates about the appropri-
Virtue is then ‘a way of doing’ and a ‘form of life’ (168). Or ateness of his intuitions.’ (Dreyfus and Dreyfus 1991,
as I said it elsewhere, in the context of thinking about health 241).
care: good health care is ‘in the how’; it is about knowing-
Against Dewey, who thinks that in difficult situations
how to deal with people and handle technologies (Coeckel-
we need deliberation and examination, Dreyfus and Drey-
bergh 2014). Of course we can reflect and deliberate. This
fus argue that of course experts should have a dialog about
may help. But if we have that kind of practical wisdom, that
their responses, but there may be no final agreement; the
kind of know-how, there is nothing in terms of virtue that is
experts have to appreciate each other’s decisions (242). And
wanted in addition or outside of that wisdom.
against Kohlberg and Habermas (and partly in agreement
with Gilligan), they argue that the highest stage of moral
5.2 Ethical expertise: Dreyfus contra Dewey
development is not about detached and abstract reasoning,
but intuitive response to the situation. They write: ‘if being
This interpretation of virtue in terms of know-how is in line
good means being able to learn from experience and use
with Dreyfus’s own phenomenology of ethical expertise, but
what one has learned so as to respond more appropriately
he takes distance from Dewey. In a paper with his brother
to the demands of others in the concrete situation, the high-
(Dreyfus and Dreyfus 1991) he repeats the account of skill
est form of ethical comportment consists in being able to
acquisition, but then draws out the implications for ethi-
stay involved and to refine one’s intuitions.’ (247). Detached
cal expertise. In so far as ethical comportment is a form of
reflection is resorted to in case of a kind of breakdown, it
expertise, the brothers Dreyfus argue, it has the same devel-
is not the rule. One could conclude that, according to the
opmental structure (236). As a child we learn moral rules,
Dreyfus brothers, skillful coping without deliberation is the
but rules and maxims are for beginners in moral develop-
‘default’, at least for experts.
ment. Ethical experts do it differently: ‘with enough expe-
Dreyfus and Dewey may actually be closer than presented
rience, the ethical expert would learn to tell the truth or
here. Both agree that the main challenge is to appropriately
lie, depending upon the situation, without appeal to rules
respond to situations. Both may also agree that emotions
and maxims’ (237). Interestingly, here the Dreyfus broth-
and intuitions play a role. Dewey also criticized abstract
ers do use Dewey to develop their account. They endorse
principles and detached reasoning. Principles are tools to
Dewey for saying that ethical comportment is spontaneous,
cope with a situation. Dreyfus could have arrived at a less
involving an immediate reaction. But then they argue that
extreme view if he recognized that, like other tools, princi-
Dewey was wrongly focused on problem solving and ignored
ples and reasoning can help us to skillfully and expertly cope
‘ongoing coping’, which is not only simple but also complex.
with ethically challenging situations, especially in a social
According to Dreyfus, as experts we can respond to complex
context. These tools may not be sufficient. And it may well
situations without deliberation (238). Against Aristotle’s
be that before, during, and after the decision, the expert can-
intellectualism (or, in Dreyfus’s view: against intellectualist
not entirely render her decision process and the knowledge
interpretations of Aristotle), the Dreyfus brothers argue that
basis of this decision process transparent. But talking with
people should not only be praised for what they intend to do
others and the use of principles may help, like other tools
but also and especially for their ‘brilliant intuitive responses’
may help. If Dreyfus would have recognized that language,
as experts (239). Virtue, then, is spontaneous: ‘We can only
concepts, etc. are also a kind of tools, he could have held the
tell if a person is courageous, for example, by seeing his
view that deliberation is perhaps not necessary, but helpful
spontaneous response in many different situations.’ (239)
as one of the tools we have. He could have arrived at the
For Dreyfus, no practical reasoning is necessary. Principles
more moderate view that moral experts at a higher stage of
can even hinder good ethical responses:
development, when faced with a problem in a social context,
‘an ethical expert when confronted with cases of “life- often masterly know how to combine various tools and types
boat morality” may have to fall back on ethical princi- of knowledge, including intuition, emotions, principles, dis-
ples. But since principles are unable to produce expert course, etc.—without being able to fully explain themselves
behavior, it should be no surprise if falling back on afterwards but with being able to have a meaningful conver-
them produces inferior responses. The resulting deci- sation with others about their decision. However, perhaps the

13
AI & SOCIETY

final aim and the highest stage is moral wisdom and virtue reveals. (.. .) Indeed, in an important sense this expe-
understood as knowing-how to do good, which may also be rience turns you into an automated device the GPS
reached without deliberation and conversation; they are nei- can use to arrive at its destination.’ (Dreyfus and Kelly
ther sufficient nor necessary. Moreover, with Dewey we can 2011, 215).
add that when we seek consensus, it is impossible to reach
While I agree with the authors on the value of skilled
a final agreement, with “final” meaning: an agreement for
activity, and even on the potentially alienating aspect of tech-
once and for all. Dreyfus could have agreed with this view.
nologies such as GPS, alienation and nihilism are certainly
But with Dewey we must appreciate, more than Dreyfus did,
not the only possibilities when it comes to contemporary
the social and collective dimension of coping and indeed of
technologies. First, the world is a lot less disenchanted than
the problems we face. There is often the practical–social
Dreyfus and Kelly suppose. Consider for instance how much
need for collective coping to reach a temporary consensus
wonder there is in contemporary science and technology,
among experts. Language and other tools may help with this.
and how romanticism and technologies are still entangled
(Coeckelbergh 2017d). More generally, it is highly doubtful
5.3 If skilled engagement and craft work are
if we find ourselves in the nihilistic situation Dreyfus and
good, what does that imply for contemporary
Kelly describe. We always find and create meaning, also in
technologies?
our daily coping with and through technologies. Second,
even if there is still alienation due to technologies, new tech-
Next to implications for a conception of virtue and moral
nologies are not only a problem; they may also be part of
development, Dreyfus’s work also contains a particular
the solution. Different technologies and uses are possible.
view of the good life—especially the good life in the con-
As postphenomenology reminds us (Ihde 1990), and as I
text of modernity. Dreyfus does not only make a descrip-
have argued using Wittgenstein (Coeckelbergh 2017a, b),
tive claim about knowledge and skill; we can also discern
technologies are ambiguous and ‘multistable’ in the sense
in this work the normative claim that skilled engagement
that they are always technologies-in-use: the same artifact
itself is something good, that particular technological prac-
has different meanings depending on use and context. This
tices—craft work especially—are ethically good in several
has implications for thinking about technology and virtue.
ways, are conducive to the good life. In line with work by
In my work on skill, technology, and virtue (Coeckelbergh
Pirsig, Sennett, Crawford, and in philosophy of technol-
2011a, 2012, 2015), I have proposed a more constructive
ogy Borgmann (see my discussion in Coeckelbergh 2015),
approach than Heidegger and Dreyfus: one which does not
Dreyfus thought that there is something special about skilled
a priori exclude the possibility that new technologies may
engagement that makes it good in itself, and that helps us
contribute to skilled activity, virtue, and good. For example,
to cope with the nihilist tendencies in modern culture. Now
when it comes to wayfinding it is true that GPS may divert
this claim—which, by itself, has a lot to say for it—is not
our attention from our surroundings, but perhaps some apps
necessarily right in its evaluation of modernity and can lead
can help us to see features of the environment that we did
to a conservative and pessimistic position with regard to
not notice (yet). There may be more modes and possibili-
new technologies. This seems the direction Dreyfus took, for
ties (Coeckelbergh 2015, 149) and more creative uses of the
example, in his book with Kelly (Dreyfus and Kelly 2011).
same device to explore different ways of wayfinding and
According to Dreyfus and Kelly, our modern (self-)under-
walking (152).
standing has left us with a meaningless world. In response
More generally, I have argued that contemporary informa-
to this nihilist condition, they argue that we must re-enchant
tion and communication technologies could, in principle,
the world by seeing that the gods are there (polytheism), that
also help us to overcome alienation and find a better relation
there is the sacred, and that there are still Greek-style heroes,
to our environment and to others. In ‘Technology as skill and
for example in sport (Dreyfus and Kelly 2011, 201). But they
activity’ (2012), I have argued that we can take seriously the
also propose craftwork and its skilled activity as a response
problem of alienation but at the same time have a Dreyfus-
to alienation and nihilism. According to the authors, we can
inspired (but also more social and Dewey-inspired) ethics
come to discern meanings that are already there (209) and
of skill that is open to the possibility that new technologies
may feel that there is a force outside our self (8). The crafts-
can also help us to overcome alienation. We can use the
man achieves ‘intimacy’ with the material (2010), whereas
concept of skill to evaluate new technologies, asking if they
modern technology such as GPS alienates us from our envi-
contribute to more or less skilled engagement. But there is
ronment. They write:
no a priori answer as to which technologies are more ‘focal’
‘to lose the sensitivities—to the landmarks, street (to use Borgmann’s term; Borgmann 1984) or more engag-
signs, wind direction, the height of the sun, the stars— ing and more conducive to skilled activity and craftsman-
all the meaningful distinctions that navigation skill ship—activities and modes of knowledge and interaction

13

which Dreyfus rightly valued. What matters then is not what
technology “is” (as is sometimes suggested in Dreyfus’s late
Heideggerian approach to new technologies), but rather what
kind of skills–activities are promoted and supported by spe-
cific technologies, by a particular technology-in-use. What
matters is whether or not the skills–activities lead to more
engagement and less alienation. (Coeckelbergh 2012).
Thus, new technologies are far more ambiguous with
regard to skilled engagement than Dreyfus suggests. Con-
sider also the work of Shannon Vallor, who has argued that
while contemporary ICTs can lead to ‘moral deskilling’
there may be also potential for ‘upskilling’ (Vallor 2015).
Recently, she has called for a technomoral virtue ethic (Val-
lor 2016). These interpretations of the relation(s) between
technology, skill, and virtue suggest that it is possible to talk
about virtue and skill in a way that is critical of contempo-
rary technologies, but also takes a more constructive and
hopeful approach: perhaps there often is, but there does not
need to be, an opposition between the cultivation of virtue
and the development of advanced technologies.
Consider again the use of electronic technologies in
health care. In ‘E-care as craftsmanship’ (2013b) I have
argued that while under modern conditions and with the use
of tele-care there are real dangers with regard to the erosion
of craftsmanship, the use of new technology is not neces-
sarily jeopardizing skillful and careful engagement with
patients and quality care. Whether such an engagement is
possible depends on whether in a specific practice and given
a specific technology e-care workers can develop the know-
how and skill to engage more intensely with those under
their care, and to cooperate with their co-workers. What
matters is how care workers work and care, when using
and working with new technology—and with each another.
Hence, craftsmanship in “e-care” is possible, for example,
when tele-care technologies are used as a complement rather
than a replacement of care workers, and when criteria and
conditions for craftsmanship are fulfilled (Coeckelbergh
2013b). Another way of putting this is, again, to say that
good health care is not about the technologies as such; good
health care is in the “how” (Coeckelbergh 2014). The chal-
lenge is to design and use technologies in such a way that
more engaged and care-full ways of doing and knowing-how
are promoted.
5.4 The social and political dimension of virtue
and craftsmanship: Dewey and MacIntyre
As I already suggested, craftsmanship also has a social
dimension (Coeckelbergh 2012). It is not only a kind of
technical skill; it is also about doing things together, shar-
ing good, and indeed the skill to work and relate to others.
But this social dimension is not always presents in views of
craftsmanship, skill, and virtue. Often the emphasis is on the
13
AI & SOCIETY
individual or personal level. This is also the case in Drey-
fus’s account of skillful coping. As I proposed, the social
dimension could be developed by bringing in a pragmatist
(or other) conception of habit and other concepts that help
us to understand the social and public dimension of skilled
engagement and virtue. Individual or personal experience
and wisdom are always linked to collective knowledge—
explicitly and implicitly. If virtue and the highest stage of
moral development are a matter of skillful coping, as Drey-
fus argued, then this skillful coping should not only be con-
ceptualized in terms of individual knowledge. Dewey’s prag-
matism or the social sciences could help Dreyfus-inspired
philosophers to further work on this problem. Another inter-
esting source may be MacIntyre, who is acknowledged by
Dreyfus but not fully used. His conception of practice as
a social and historically situated activity that enables par-
ticipants to attain goods internal to the practice and achieve
excellence and virtue (MacIntyre 1984) might have enabled
Dreyfus to say more about the social dimension of skillful
coping and virtue, in particular about the links between skill-
ful coping, virtue, and practice. It may certainly provide a
source of inspiration to those philosophers working on virtue
and technology who are interested in developing the social
dimension of their work. (Note also that MacIntyre, Dewey,
and the later Wittgenstein are in agreement when it comes
to rejecting the fact/value distinction, which may also help
to conceptualize a view of virtue that links it firmly to its
social and practical contexts. However, I will not further
elaborate this point here.)
Moreover, given the often-lacking attention for the social
dimension of virtue in modern theories of virtue, virtue
approaches are in danger of ignoring the political dimen-
sion—including potential political problems that are often
ascribed to thinking in terms of skill, craft, and virtue. Since
modern virtue ethicists tend to be focused on individual
moral development or (more local) practices, they are often
blind to the wider and often antagonistic social playing field,
within which individual cultivation of character is set. Or far
worse, they do not ask the political question at all.
This is also a problem for a Dreyfus-inspired thinking
about technology, to the extent that it comes in the shape of
a virtue ethics focused on individual skillful coping. Such
a focus leaves too much room for interpretation concerning
the political implications. Reynolds (2006) has argued that
there is a danger that Dreyfus’s phenomenology devolves
into a conservative communitarianism that does not recog-
nize ‘the inevitability and vitality of social conflict.’ (555).
Whether or not this is true, it raises the issue of the political
implications of Dreyfus’s phenomenology. Dreyfus seems
to have inherited Heidegger’s positive valuation of the kind
of knowledge that is involved in craft work. But did he also
inherit Heidegger’s conservatism or worse, his Nazism?
Does a positive valuation of craft work or a tendency
AI & SOCIETY
towards virtue ethics necessarily lead to conservatism? Is
it communitarian, and then what kind of communitarian-
ism? Is communitarianism necessarily conservative? Or are
there other possibilities? And is there an unbridgeable gap
between phenomenology and critical theory, or could Drey-
fus’s phenomenology be combined with some concerns and
methods from critical theory?
Reading Dreyfus’s work I do not see indications in the
direction of conservatism, let alone Nazism, and I do not see
why his thinking would necessarily lead into that direction.
There is conservatism regarding technology, perhaps, but
this does not necessarily imply social or political conserva-
tivism. However, these questions deserve further discussion,
also in the light of contemporary debates about Heidegger’s
political sympathies. Since, as I have argued, Dreyfus’s
thinking has insufficiently engaged with the problem of the
social, and since it has never been fully developed in terms
of its social and political implications, it remains somewhat
vulnerable to typical anti-Heideggerian criticisms. There-
fore, if philosophers of technology want to use Dreyfus’s
thinking, they need to say more about the relations between
skill and the social, and explicitly reflect on the normative
and political directions this may take them. In the previous
pages I have proposed some ways in which this can be done
by staying relatively close to Dreyfus’s own sources (rather
than resorting to critical theory, for instance). In particular,
I have suggested that using and combining Wittgenstein and
Dewey (and maybe also MacIntyre) has a lot of potential to
say more about the social in relation to the phenomenology
of skillful coping.
More generally, more work is needed to develop these
thoughts about how Wittgenstein, Dewey, and Dreyfus can
be combined to (re)conceptualize and discuss virtue and
its social dimension. This project requires, among other
things, further engagement with Wittgenstein’s work and
its interpreters and a more elaborate discussion of both con-
vergences and tensions between these thinkers. Here I have
limited myself to offering some signposts in this direction,
and to exploring some potential implications for philosophy
of technology.
6 Conclusion
For thinking about technology, Dreyfus’s account of skilled
coping is and remains a valuable source, especially for
those who are, like Dreyfus, interested in using Heidegger,
Merleau-Ponty, and the later Wittgenstein to better under-
stand the knowledge and experience involved in the use of
tools, in skillful coping with and through technologies. How-
ever, I have also outlined and discussed some challenges
that come with Dreyfus’s approach: problems with regard
to his assumptions about technology, his underdeveloped
conceptions of the social dimension of skillful coping, and
uncertainty with regard to the ethical and political implica-
tions of this view, including questions regarding the relation
between virtue and skill. I have indicated some potential
avenues for addressing these problems. In general, my rec-
ommendation to those inspired by Dreyfus in philosophy
of technology is to take more distance from the later Hei-
degger’s conception of technology and to do more with
Wittgenstein and Dewey than Dreyfus did to deal with these
challenges. I have indicated that there is also a lot of room
for further development in the direction of virtue theory, for
example by combining Dreyfus and Dewey or by engaging
with the work of MacIntyre. Finally, so far authors working
in the field of postphenomenology have not been very inter-
ested in Dreyfus’s work (and vice versa); perhaps since the
former rejects the latter as too Heideggerian. This neglect is
to be regretted since Dreyfus’s work, in particular its account
of skillful coping, has a lot to offer to anyone interested
in the phenomenology of technology use. By shifting the
emphasis from the material artifact as such to its use and
the skills and knowledge involved in that use, contempo-
rary philosophers of technology in postphenomenology and
related approaches could work towards a more relational and
non-dualistic thinking about the relations between humans
and technology—an aim Ihde may sympathize with.
More generally, Dreyfus’s work constitutes a rich poten-
tial source of inspiration for philosophers of technology
today, not only in the area of philosophy of AI but—with the
account of skillful coping—also in other areas of thinking
about technology. It is especially of interest to philosophers
who aim to be sensitive to the phenomenology of living with
technologies, to human experience with technologies. This
attention to technology in the lifeworld also requires us to
adopt, in contrast to Dreyfus and Heidegger, a view of new
technologies that is more open and constructive and that
goes beyond embodied tool use, without, however, dismiss-
ing the concerns about skill, alienation and virtue Dreyfus
and others rightly raised. This could help philosophy of
technology to further develop thinking about the phenom-
enological and ethical implications of AI: not in the abstract,
but in use, as it is becoming part of our lifeworld and indeed
of our skilled practices and our coping.
But Dreyfus’s work also remains relevant to contempo-
rary philosophy in general. Its main message is especially
relevant and urgent when, and to the extent that, this philoso-
phy is too mesmerized by formal and theoretical knowledge
and by abstract arguments and principles, to the neglect of
other, less explicit types of knowledge involved our everyday
living and coping. Together with other key twentieth-century
philosophers (and indeed contemporary cognitive science),
Dreyfus successfully criticized the idea that we can sepa-
rate mind and thinking from bodies and living people. If
that were possible, we would not miss Dreyfus’s voice here
13

and now. The writings and other memory technologies we
have would do the job of representing and preserving his
thinking. But, as Dreyfus taught generations of philosophers
by means of his teaching and (in my case) writings, nei-
ther thinking nor human beings can be reduced to what can
be represented. Thinking needs to be embodied and lived.
And, therefore—if not for other reasons—Hubert Dreyfus
is greatly missed.
Acknowledgment Open access funding provided by University of
Vienna.
Open Access This article is distributed under the terms of the Crea-
tive Commons Attribution 4.0 International License (http://creativeco
mmons.org/licenses/by/4.0/), which permits unrestricted use, distribu-
tion, and reproduction in any medium, provided you give appropriate
credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made.
References
Andler D (2000) Context and background: Dreyfus and cognitive sci-
ence. In: Wrathall M (ed) Heidegger, coping and cognitive sci-
ence. MIT Press, Cambridge, pp 137–159
Annas J (1993) The morality of happiness. Oxford University Press,
Oxford
Annas J (1995) Virtue as a skill. Int J Philos Stud 3(2):227–243. https
://doi.org/10.1080/09672559508570812
Annas J (2007) Virtue ethics. In: Copp D (ed) The Oxford handbook
of ethical theory. Oxford University Press, Oxford
Borgmann A (1984) Technology and the character of contemporary
life: a philosophical inquiry. The University of Chicago Press,
Chicago
Breivik G (2007) Skillful coping in everyday life and in sport: a critical
examination of the views of Heidegger and Dreyfus. J Philos Sport
34(2):116–134. https://doi.org/10.1080/00948705.2007.9714716
Carman T (2008) On being social: a reply to Olafson. Inquiry
37(2):203–223
Coeckelbergh M (2011a) Environmental virtue: motivation, skill, and
(in)formation technology. J Environ Philos 8(2):141–170
Coeckelbergh M (2011b) Can we trust robots? Ethics Inf Technol
14:53–60
Coeckelbergh M (2012) Technology as skill and activity: revisiting the
problem of alienation. Techné 16(3):208–230
Coeckelbergh M (2013a) Human being@ risk: enhancement, technol-
ogy, and the evaluation of vulnerability transformations. Springer,
Dordrecht
Coeckelbergh M (2013b) E-care as craftsmanship: virtuous work,
skilled engagement, and information technology in health care.
Med Healthc Philos 16(4):807–816
Coeckelbergh M (2014) Good healthcare is in the how: the quality of
care, the role of machines, and the need for new skills. In: Coeck-
elbergh M, van Rysewyk SP, Pontier M (eds) Machine medical
ethics. Springer, Berlin, pp 33–47
Coeckelbergh M (2015) Environmental skill: motivation, knowledge,
and the possibility of a non-romantic environmental ethics. Rout-
ledge, New York
Coeckelbergh M (2016) Responsibility and the moral phenomenology
of using self-driving cars. Appl Artif Intell 30(8):748–757
13
AI & SOCIETY
Coeckelbergh M (2017a) Using words and things: language and phi-
losophy of technology. Routledge, New York
Coeckelbergh M (2017b) Technology games: using Wittgenstein for
understanding and evaluating technology. Sci Eng Ethics https://
doi.org/10.1007/s11948-017-9953-8
Coeckelbergh M (2017c) Language and technology: maps, bridges, and
pathways. AI Soc 32(2):175–189
Coeckelbergh M (2017d) New romantic cyborgs: romanticism, infor-
mation technology, and the end of the machine. MIT Press,
Cambridge
Coeckelbergh M, Funk M (2018) Wittgenstein as a philosopher
of technology: tool use, forms of life, technique, and a tran-
scendentalargument. Hum Stud https://doi.org/10.1007/s1074
6-017-9452-6(online first)
Collins H (1992) Hubert L. Dreyfus, forms of life, and a simple test
for machine intelligence. Soc Stud Sci 22:726–739. http://www.
jstor.org/stable/285462
Collins H (2013) The core of expertise. Phenomenol Cognit Sci
12:399–416. https://doi.org/10.1007/s11097-012-9277-8
Crawford MB (2009) Shop class as soulcraft: an inquiry into the
value of work. Penguin Press, New York
Dewey J (1922) Human nature and conduct. Allen & Unwin, London
Dewey J (1929) Experience and nature. George Allen & Unwin,
London
Dreyfus HL (1967) Why computers must have bodies in order to be
intelligent. Rev Metaphys 21(1):13–32
Dreyfus HL (1972) What computers can’t do. the limits of artificial
intelligence. MIT Press, New York
Dreyfus HL (1975) The priority of the world to my world: Hei-
degger’s answer to Husserl (and Sartre). Man World 8:121–130
Dreyfus HL (1980b) Holism and hermeneutics. Rev Metaphys
34(1):3–23
Dreyfus HL (1991) Towards a phenomenology of ethical expertise.
Hum Stud 14:229–250
Dreyfus HL (1992) Response to Collins, artificial experts. Soc Stud
Sci 22(4):717–726
Dreyfus HL (1999) The primacy of phenomenology over logical
analysis. Philos Top 27(2):3–24
Dreyfus HL (2001a) Phenomenological description versus rational
reconstruction. Rev Int Philos 2(216):181–196
Dreyfus HL (2001b) On the internet (revised 2nd edition), 2009.
Routledge, London
Dreyfus HL (2002a) Intelligence without representation—Merleau-
Ponty’s critique of mental representation. Phenomenol Cognit
Sci 1:367–383. https://doi.org/10.1023/A:1021351606209
Dreyfus HL (2002b) Refocusing the question: can there be skillful
coping without propositional representations or brain represen-
tations? Phenomenol Cognit Sci 1:413–425
Dreyfus HL (2005) Overcoming the myth of the mental: how philos-
ophers can profit from the phenomenon of everyday expertise.
Proc Addresses Am Philos Assoc 79(2):pp 47–65
Dreyfus HL (2007a) The return of the myth of the mental. Inquiry
50(4):352–365. https://doi.org/10.1080/00201740701489245
Dreyfus HL (2007b) Response to McDowell. Inquiry 50(4):371–377.
https://doi.org/10.1080/00201740701489401
Dreyfus HL (2012) Introductory essay: the mystery of the back-
ground qua background. In: Radman Z (ed) Knowing without
thinking. Palgrave Macmillan, Basingstoke, pp 1–10
Dreyfus HL (2014) Skillful coping: essays on the phenomenology
of everyday perception and action. In: Wrathall M (ed) Oxford
University Press, Oxford
Dreyfus S, Dreyfus HL (1980a) A five-stage model of the mental
activities involved in directed skill acquisition. Operations
Research Centre, University of California, Berkeley
Dreyfus HL, Kelly SE (2011) All things shining. Reading theWestern
classics to find meaning in a secular age. Free Press, New York
AI & SOCIETY
Heidegger M (1927) Being and time: a translation of Sein und Zeit,
Stambaugh J (trans, 1996). State University of New York,
Albany
Heidegger M (1977) The question concerning technology. In: Lovitt
W (trans) The question concerning technology and other essays.
Harper & Row, New York, pp 3–35
Ihde D (1990) Technology and the lifeworld: from garden to earth.
Indiana University Press, Bloomington
MacIntyre A (1984) After virtue: a study in moral theory, 3rd edn.
2007. University of Notre Dame Press, Notre Dame
McDowell J (2007) Response to Dreyfus. Inquiry 50(4):366–370. https
://doi.org/10.1080/00201740701489351
McManus D (2007) Rules, regression and the ‘Background’: Dreyfus,
Heidegger and McDowell. Eur J Philos 16(3):432–458. https://
doi.org/10.1111/j.1468-0378.2007.00255.x
McPherson I (2005) Reflexive learning: stages towards wisdom with
Dreyfus. Educ Philos Theory 37(5):705–718. https://doi.org/10.
1111/j.1469-5812.2005.00152.x
Merleau-Ponty M (1945) Phenomenology of perception. Smith C
(trans, 1962). Routledge, New York
Olafson FA (2008) Heidegger à la Wittgenstein or ‘coping’ with Pro-
fessor Dreyfus. Inquiry 37(1):45–64
Reynolds J (2006) Dreyfus and deleuze on, coping, and trauma in skill
acquisition. Int J Philos Stud 14(4):539–559
Rietveld E (2010) McDowell and Dreyfus on unreflective action.
Inquiry 53(2):183–207. https ://doi.org/10.1080/00201 74100
3612203
Selinger E, Crease RP (2003) Dreyfus on expertise: the limits of phe-
nomenological analysis. Cont Philos Rev 35:245–279
Stichter M (2007) Ethical expertise: the skill model of virtue. Ethi-
cal Theory Moral Pract 10:183–194. http://www.jstor.org/stabl
e/40602522
Vallor S (2015) Moral deskilling and upskilling in a new machine age:
reflections on the ambiguous future of character. Philos Technol
28:107–124
Vallor S (2016) Technology and the virtues: a philosophical guide to a
future worth wanting. Oxford University Press, New York
Verbeek P (2005) What things do: philosophical reflections on technol-
ogy, agency, and design. The Pennsylvania State University Press,
University Park
Winner L (1986) The whale and the reactor: a search for limits in an
age of high technology. The University of Chicago Press, Chicago
Wittgenstein L (1969) On Certainty. In: Anscombe GEM, von Wright
GH (eds) Basil Blackwell, Oxford(trans: Paul D, Anscombe
GEM)
Wittgenstein L,. (trans: Anscombe E, Hacker PMS, Schulte J
(1953/2009) Philosophical investigations, 4th edn. Wiley-Black-
well, Malden
13